Skip to content
Agentmart
Submit a tool

update-agents-md

Generate or refresh a repo's AGENTS.md (the source-of-truth agent context file) and its one-line CLAUDE.md pointer.

Skill by therapys

IndexedMIT★ 8
Write a review

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/therapys-update-agents-md

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Generate or refresh a repo's AGENTS.md (the source-of-truth agent context file) and its one-line CLAUDE.md pointer. Use when the user says 'update the claude.md', 'update AGENTS.md', 'analyze this project and update the context file', 'write an AGENTS.md', or after a stack/structure change makes the current file stale.

Use cases: writing, agent-building

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/update-agents-md

What's inside

Skill name
update-agents-md
Description
Generate or refresh a repo's AGENTS.md (the source-of-truth agent context file) and its one-line CLAUDE.md pointer. Use when the user says 'update the claude.md', 'update AGENTS.md', 'analyze this project and update the context file', 'write an AGENTS.md', or after a stack/structure change makes the current file stale.
License
MIT
Instructions
41 lines in SKILL.md · view on GitHub

Files (1)

  • SKILL.md

SKILL.md (first lines)

# Update AGENTS.md

This skill follows a two-file convention: **`AGENTS.md` is the single source of truth**, and **`CLAUDE.md` is a one-line pointer** containing exactly `@AGENTS.md`. pi reads `AGENTS.md` directly; Claude Code reads `CLAUDE.md`, which imports it. Maintain AGENTS.md — never duplicate content across the two files.

## Steps

1. **Ground yourself in the repo — don't invent:**
   - *Stack:* read `package.json` (dependencies + `scripts`), the lockfile (bun/pnpm/npm — tells you the package manager), framework config, `tsconfig`, and linter/formatter config.
   - *Layout:* list the real top-level and `src/` dirs; open a few to learn what each actually holds.
   - *Conventions:* infer from existing code — import alias, state/data libraries, auth, testing, formatting — and cite real paths.
2. **Write/refresh `AGENTS.md`** using the skeleton below. Keep it dense and specific: every line should save an agent a lookup. No filler, no restating generic framework docs.
3. **Ensure `CLAUDE.md` is exactly `@AGENTS.md`** (create or replace). If CLAUDE.md already holds real content and there's no AGENTS.md, migrate that content into AGENTS.md first, then reduce CLAUDE.md to the one-line pointer.
4. **Preserve tool-managed blocks verbatim** — e.g. `<!-- BEGIN:nextjs-agent-rules -->…<!-- END:nextjs-agent-rules -->` is rewritten by `next dev`. Keep it in place and commit it with your changes (removing it only re-creates an uncommitted diff).

## Section skeleton (AGENTS.md)

```
## Stack
- framework + versions, package manager, db/orm, auth, styling, state, lint/format — one bullet each, real versions

## Repo Layout
- `dir/` — what lives here and the one rule an agent must know (cite real paths)

## Runtime Patterns
- how requests / data / sessions actually flow; the non-obvious wiring

## Conventions
- import alias, where to add X before creating new, naming, do / don't

## Agent Behavior
- repo-specific rules: what to run before committing, what never to touch, gotchas
```

## Guardrails

- Match the terseness of the repo's existing AGENTS.md files (~40–110 lines). If it's growing long, cut — link to code instead of explaining it.
- Never document a path or pattern you haven't verified in the tree (e.g. don't reference an old `sessions.ts` if the code now lives in `src/lib/sessions/`).
- No license boilerplate, no table of contents, no "this file provides guidance" preamble.
- Leave committing/PRs to the `commit` / `ship` skills.

From therapys/dotagents, MIT.

More from therapys/dotagents

Signals

Stars
8
Forks
0
Open issues
0
Last commit (this folder)
Aug 28, 2026 (2 months ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from therapys

Cite this page

update-agents-md skill. Agent Mart. https://agentmart.dev/skills/therapys-update-agents-md (last updated 2026-10-24)